Techniques for surfacing relevant reviews are disclosed. In some embodiments, one or more reviews are annotated with one or more applicable tags. Annotations of reviews associated with a subject of interest are aggregated to identify an applicable insight into the subject of interest, and the identified insight is published on a page associated with the subject of interest, wherein the published insight includes an option to navigate to a subset of reviews associated with the subject of interest that are relevant to the insight.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method, comprising: annotating one or more reviews with one or more applicable tags; aggregating annotations of reviews associated with a subject of interest to identify an applicable insight into the subject of interest; and publishing the identified insight determined from reviews of the subject of interest on a page associated with the subject of interest, wherein the published insight comprises a selectable hyperlink to navigate to pages of one or more other subjects of interest on which the same insight is published.
A method for surfacing relevant reviews involves three steps. First, one or more reviews are automatically annotated with applicable tags. Second, annotations of reviews related to a specific topic are aggregated to determine a key insight about that topic. Third, this insight, derived from the aggregated reviews, is published on a webpage dedicated to that topic, including a hyperlink that allows users to navigate to other pages showing the same insight on different subjects.
2. The method of claim 1 , wherein annotating one or more reviews with one or more applicable tags comprises annotating a review based on one or more of reviewer data, review content, and review rating.
The review annotation method from the previous description automatically assigns tags to reviews based on reviewer information (e.g., demographics), the content of the review itself (e.g., keywords), and the rating provided by the reviewer (e.g., star rating). This multifaceted analysis improves tag accuracy.
3. The method of claim 1 , wherein annotating one or more reviews with one or more applicable tags comprises annotating a review with a tag if the review includes at least a threshold number of terms associated with the tag.
The review annotation method from the initial description tags reviews based on the number of times specific keywords appear within the review. If the number of these keywords exceeds a predefined threshold, the review is automatically tagged accordingly, indicating a strong association between the review and the tag's topic.
4. The method of claim 1 , wherein annotating one or more reviews with one or more applicable tags comprises annotating one or more reviews posted by a reviewer with a tag associated with a special interest group or demographic to which the reviewer is determined to belong.
The review annotation method from the initial description automatically tags reviews written by users associated with a specific interest group or demographic. If a reviewer is identified as belonging to such a group, their reviews are tagged with labels relevant to that group, reflecting their potential biases or specialized knowledge.
5. The method of claim 1 , wherein annotating one or more reviews with one or more applicable tags comprises annotating a review with a tag based on a positive or negative rating provided with the review.
The review annotation method from the initial description uses the review's rating (positive or negative) to determine the appropriate tags. A positive review gets associated with positive sentiment tags, and a negative review gets associated with negative sentiment tags, reflecting the reviewer's overall opinion.
6. The method of claim 1 , wherein aggregating annotations of reviews associated with the subject of interest comprises determining a set of associated tags and a number of reviews that are annotated with each different tag.
The review aggregation method from the initial description compiles all the tags assigned to reviews about a specific topic. It then counts the number of reviews associated with each unique tag. This results in a set of tags ranked by the frequency with which they appear in reviews about that topic.
7. The method of claim 6 , wherein identifying an applicable insight into the subject of interest comprises determining that a number of reviews annotated with a prescribed tag associated with the insight satisfies a threshold.
The insight identification method described in the previous review aggregation process determines if the number of reviews with a specific tag, associated with the insight, exceeds a predefined minimum. If the tag appears frequently enough (above the threshold), the corresponding insight is considered valid and is published.
8. The method of claim 1 , wherein publishing the identified insight comprises publishing the insight as a badge.
The method described initially publishes the identified insight by visually representing it as a badge on the topic's webpage. The badge provides a quick and prominent summary of the most relevant information extracted from the reviews.
9. A system, comprising: an annotation engine comprising a processor configured to annotate one or more reviews with one or more applicable tags; and an aggregation engine comprising a processor configured to: aggregate annotations of reviews associated with a subject of interest to identify an applicable insight into the subject of interest; and publish the identified insight determined from reviews of the subject of interest on a page associated with the subject of interest, wherein the published insight comprises a selectable hyperlink to navigate to pages of one or more other subjects of interest on which the same insight is published.
A system for surfacing relevant reviews has two components: an annotation engine and an aggregation engine, both using processors. The annotation engine automatically tags reviews. The aggregation engine then analyzes these tags to find key insights. Finally, the aggregation engine publishes these insights on relevant topic pages, adding a hyperlink that connects to other pages where the same insight is shown for other subjects.
10. The system of claim 9 , wherein to annotate one or more reviews with one or more applicable tags comprises to annotate a review based on one or more of associated reviewer data, review content, and review rating.
The review annotation functionality within the system from the previous description automatically assigns tags to reviews based on reviewer information (e.g., demographics), the content of the review itself (e.g., keywords), and the rating provided by the reviewer (e.g., star rating). This multifaceted analysis improves tag accuracy.
11. The system of claim 9 , wherein to annotate one or more reviews with one or more applicable tags comprises to annotate a review with a tag if the review includes at least a threshold number of terms associated with the tag.
The review annotation functionality within the system from the previous description tags reviews based on the number of times specific keywords appear within the review. If the number of these keywords exceeds a predefined threshold, the review is automatically tagged accordingly, indicating a strong association between the review and the tag's topic.
12. The system of claim 9 , wherein to annotate one or more reviews with one or more applicable tags comprises to annotate one or more reviews posted by a reviewer with a tag associated with a special interest group or demographic to which the reviewer is determined to belong.
The review annotation functionality within the system from the previous description automatically tags reviews written by users associated with a specific interest group or demographic. If a reviewer is identified as belonging to such a group, their reviews are tagged with labels relevant to that group, reflecting their potential biases or specialized knowledge.
13. The system of claim 9 , wherein to annotate one or more reviews with one or more applicable tags comprises to annotate a review with a tag based on a positive or negative rating provided with the review.
The review annotation functionality within the system from the previous description uses the review's rating (positive or negative) to determine the appropriate tags. A positive review gets associated with positive sentiment tags, and a negative review gets associated with negative sentiment tags, reflecting the reviewer's overall opinion.
14. The system of claim 9 , wherein to aggregate annotations of reviews associated with the subject of interest comprises to determine a set of associated tags and a number of reviews that are annotated with each different tag.
The review aggregation functionality within the system from the initial description compiles all the tags assigned to reviews about a specific topic. It then counts the number of reviews associated with each unique tag. This results in a set of tags ranked by the frequency with which they appear in reviews about that topic.
15. The system of claim 14 , wherein to identify an applicable insight into the subject of interest comprises to determine that a number of reviews annotated with a prescribed tag associated with the insight satisfies a threshold.
The insight identification functionality within the system described in the previous review aggregation process determines if the number of reviews with a specific tag, associated with the insight, exceeds a predefined minimum. If the tag appears frequently enough (above the threshold), the corresponding insight is considered valid and is published.
16. The system of claim 9 , wherein to publish the identified insight comprises to publish the insight as a badge.
The system described initially publishes the identified insight by visually representing it as a badge on the topic's webpage. The badge provides a quick and prominent summary of the most relevant information extracted from the reviews.
17. The system of claim 9 , wherein the aggregation engine comprises the annotation engine.
In the system described initially for surfacing reviews, the annotation engine, responsible for tagging reviews, is incorporated directly into the aggregation engine, which identifies and publishes insights. This combines tag generation and insight extraction within a single functional unit.
18. A computer program product embodied in a non-transitory computer readable storage medium and comprising computer instructions for: annotating one or more reviews with one or more applicable tags; aggregating annotations of reviews associated with a subject of interest to identify an applicable insight into the subject of interest; and publishing the identified insight determined from reviews of the subject of interest on a page associated with the subject of interest, wherein the published insight comprises a selectable hyperlink to navigate to pages of one or more other subjects of interest on which the same insight is published.
A software program stored on a computer-readable medium automates surfacing relevant reviews. The program first annotates reviews with tags. Next, it aggregates these tags to find significant insights about a specific topic. Finally, it displays the derived insight on the relevant topic page, with a hyperlink that allows users to navigate to other pages showing the same insight on different subjects.
19. The method of claim 1 , wherein the published insight comprises a selectable hyperlink to navigate to a subset of reviews associated with the subject of interest that is relevant to the insight.
In the initial review surfacing method, the published insight contains a link that redirects users to a filtered list of reviews specifically relevant to the displayed insight on that topic, allowing for deeper exploration.
20. The system of claim 9 , wherein the published insight comprises a selectable hyperlink to navigate to a subset of reviews associated with the subject of interest that is relevant to the insight.
In the system described initially for surfacing reviews, the published insight contains a link that redirects users to a filtered list of reviews specifically relevant to the displayed insight on that topic, allowing for deeper exploration.
21. The computer program product of claim 18 , wherein annotating one or more reviews with one or more applicable tags comprises annotating a review based on one or more of reviewer data, review content, and review rating.
The review annotation functionality within the computer program product described initially automatically assigns tags to reviews based on reviewer information (e.g., demographics), the content of the review itself (e.g., keywords), and the rating provided by the reviewer (e.g., star rating). This multifaceted analysis improves tag accuracy.
22. The computer program product of claim 18 , wherein annotating one or more reviews with one or more applicable tags comprises annotating a review with a tag if the review includes at least a threshold number of terms associated with the tag.
The review annotation functionality within the computer program product described initially tags reviews based on the number of times specific keywords appear within the review. If the number of these keywords exceeds a predefined threshold, the review is automatically tagged accordingly, indicating a strong association between the review and the tag's topic.
23. The computer program product of claim 18 , wherein annotating one or more reviews with one or more applicable tags comprises annotating one or more reviews posted by a reviewer with a tag associated with a special interest group or demographic to which the reviewer is determined to belong.
The review annotation functionality within the computer program product described initially automatically tags reviews written by users associated with a specific interest group or demographic. If a reviewer is identified as belonging to such a group, their reviews are tagged with labels relevant to that group, reflecting their potential biases or specialized knowledge.
24. The computer program product of claim 18 , wherein annotating one or more reviews with one or more applicable tags comprises annotating a review with a tag based on a positive or negative rating provided with the review.
The review annotation functionality within the computer program product described initially uses the review's rating (positive or negative) to determine the appropriate tags. A positive review gets associated with positive sentiment tags, and a negative review gets associated with negative sentiment tags, reflecting the reviewer's overall opinion.
25. The computer program product of claim 18 , wherein aggregating annotations of reviews associated with the subject of interest comprises determining a set of associated tags and a number of reviews that are annotated with each different tag.
The review aggregation functionality within the computer program product described initially compiles all the tags assigned to reviews about a specific topic. It then counts the number of reviews associated with each unique tag. This results in a set of tags ranked by the frequency with which they appear in reviews about that topic.
26. The computer program product of claim 25 , wherein identifying an applicable insight into the subject of interest comprises determining that a number of reviews annotated with a prescribed tag associated with the insight satisfies a threshold.
The insight identification functionality within the computer program product described in the previous review aggregation process determines if the number of reviews with a specific tag, associated with the insight, exceeds a predefined minimum. If the tag appears frequently enough (above the threshold), the corresponding insight is considered valid and is published.
27. The computer program product of claim 18 , wherein publishing the identified insight comprises publishing the insight as a badge.
The computer program product described initially publishes the identified insight by visually representing it as a badge on the topic's webpage. The badge provides a quick and prominent summary of the most relevant information extracted from the reviews.
28. The computer program product of claim 18 , wherein the published insight comprises a selectable hyperlink to navigate to a subset of reviews associated with the subject of interest that is relevant to the insight.
In the computer program product described initially for surfacing reviews, the published insight contains a link that redirects users to a filtered list of reviews specifically relevant to the displayed insight on that topic, allowing for deeper exploration.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
March 4, 2013
April 11, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.